The present invention, in some embodiments thereof, relates to communication network design and, more specifically, but not exclusively, to traffic usage optimization within communication networks.
In advanced distributed networks, the amount of data to be transferred within the network may exceed the bandwidth capacity resources of the network infrastructure. Possible outcomes of traffic overload are increased latency, network congestion, loss of data, and failures.
Flow control and traffic engineering methods may be used to overcome such problems. These methods are applied online during network operation, and respond to the current state of the network, or to the near predicted future. As such, they allow preventing network congestion in the presence of short bursts of data generation, but are ineffective when the average data generation rate exceeds network capacity.
During the network design phase, utilities performing traffic reduction may be located within the network. These utilities capture the massive amount of data originated by end-points, and reduce the data volume sent onwards, resulting in significantly lower volume of data forwarded from the utilities onwards than the volume which was originally sent up to the utilities. Prominent examples of traffic reduction usage may be found in active management networks such as the power smart grid and in hierarchical large scale distributed caches across the internet.